Professional Documents
Culture Documents
Textbook Modelling With Ordinary Differential Equations Dreyer Ebook All Chapter PDF
Textbook Modelling With Ordinary Differential Equations Dreyer Ebook All Chapter PDF
Equations Dreyer
Visit to download the full and correct content document:
https://textbookfull.com/product/modelling-with-ordinary-differential-equations-dreyer/
More products digital (pdf, epub, mobi) instant
download maybe you interests ...
https://textbookfull.com/product/modelling-with-ordinary-
differential-equations-a-comprehensive-approach-chapman-hall-crc-
numerical-analysis-and-scientific-computing-alfio-borzi/
https://textbookfull.com/product/ordinary-differential-equations-
and-boundary-value-problems-volume-i-advanced-ordinary-
differential-equations-1st-edition-john-r-graef/
https://textbookfull.com/product/from-ordinary-to-partial-
differential-equations-1st-edition-giampiero-esposito-auth/
https://textbookfull.com/product/lectures-problems-and-solutions-
for-ordinary-differential-equations-second-edition-yuefan-deng/
Ordinary differential equations : an introduction to
the fundamentals Second Edition. Edition Howell
https://textbookfull.com/product/ordinary-differential-equations-
an-introduction-to-the-fundamentals-second-edition-edition-
howell/
https://textbookfull.com/product/methods-of-mathematical-
modelling-fractional-differential-equations-1st-edition-harendra-
singh/
https://textbookfull.com/product/ordinary-differential-equations-
principles-and-applications-cambridge-iisc-series-1st-edition-a-
k-nandakumaran/
https://textbookfull.com/product/handbook-of-ordinary-
differential-equations-exact-solutions-methods-and-problems-
third-edition-andrei-d-polyanin/
https://textbookfull.com/product/differential-equations-with-
matlab-2nd-edition-brian-r-hunt/
Modelling with
Ordinary Differential
Equations
Modelling with
Ordinary Differential
Equations
T.P. Dreyer
Department o f Applied Mathematics
University of Stellenbosch
Stellenbosch, South Africa
CRC Press
Taylor & Francis Group
Boca Raton London New York
Dreyer, T. P.
Modelling with ordinary differential equations / T. P. Dreyer.
p. cm.
Includes bibliographical references and index.
ISBN 0-8493-8636-5
1. Mathematical models. 2. Differential equations. I. Title.
QA401.D74 1993
511'.— dc20 93-9636
CIP
This book represents information obtained from authentic and highly regarded sources.
Reprinted material is quoted with permission, and sources are indicated. A wide variety of
references are listed. Every reasonable effort has been made to give reliable data and informa
tion, but the author and the publisher cannot assume responsibility for the validity o f all
materials or for the consequences of their use.
Neither this book nor any part may be reproduced or transmitted in any form or by any
means, electronic or mechanical, including photocopying, microfilming, and recording, or by
any information storage and retrieval system, without permission in writing from the pub
lisher.
1 In tro d u ctio n 1
1.1 M athem atical M o d ellin g ............................................................... 1
1.2 Boundary Value P ro b le m s ............................................................ 3
1.3 Direction F ie ld s................................................................................ 7
1.4 Finding Z e r o s ................................................................................... 13
1.5 Exercises ......................................................................................... 20
3 N u m erical M eth o d s 99
3.1 In tro d u c tio n ....................................................................................... 99
3/2 Existence T h e o re m .......................................................................... 100
3.3 Euler A l g o r i t h m ............................................................................. 102
3.4 Error Analysis ............................................................................... 105
3.5 R unge-K utta A lg o rith m ............................................................... 112
3.6 F ruitflies............................................................................................. 116
3.7 Exercises ......................................................................................... 121
3.8 P r o j e c t s ............................................................................................. 125
3.9 M athem atical B ack grou nd............................................................ 127
9 A n sw ers 267
The book is not a reference manual and does not pretend to cover com
pletely either one of the fields of modelling or ordinary differential equa
tions. It is prim arily meant as an elementary text to be used in a lecture
room. It can, however, also be read as a self-study program since each
model is developed from first principles, examples are provided in the
text to clarify new definitions, and carefully chosen exercises are given
in each section to strengthen the understanding of the m aterial and the
skill to construct models. Supplem entary references are also given at
the end of each section to stim ulate further reading.
The first version of this book in the form of classroom notes was w ritten
in 1972. As the m aterial was tested in the classroom over two decades,
some m aterial was discarded, new models were added, and the selection
of exercises and projects was refined. This book is the end result of a
m arriage between the academic goals set by the teacher and the feedback
through questionnaires and examinations of the students. The favorable
comments of ex-students working as applied m athem aticians in industry
indicate th at the choice of m aterial and the level of sophistication in the
book are satisfactory.
A part from the exercises, there is a section on projects at the end of each
chapter. These are mainly intended for the com puter-literate reader,
and can also be used in tutorials.
The book consists of seven chapters. In the first chapter some basic
notions are introduced. The second chapter is a diverse collection of
real-life situations where a first order differential equation pops up in
the m athem atical modelling. Each problem was chosen with a specific
purpose in mind. In general it is shown th at there is much more to these
m athem atical models than merely solving differential equations; on the
contrary, in some cases it is not even necessary to solve the differential
equation.
Through the years many people have contibuted to the creation of this
book. The discussions with my colleagues were very helpful, in particu
lar the ideas of G erhard Geldenhuys and Philip Fourie. The suggestion
to write this book came from Alan Jeffrey of the University of Newcastle
on Tyne. His interest and comments are deeply appreciated. The en
couragement of Navin Sullivan in London carried the project through.
Finally, w ithout the help of Hester Uys and Jan van Vuuren on the
computer, and Christelle Goldie with the sketches, there would not be
a book at all.
Although this book may look like a jum ble of disconnected applications,
be warned - there is m ethod in the madness, and you may learn more
than you thought you would. So start right away and enjoy the book.
T P Dreyer
1
In trod u ction
1
2 C hapter 1 Introduction
Identification
C onstruction
In te rp re ta tio n
Im plem entation
° n ^ 2 ^ + • • • + + a ° ( x ) y =
4 Chapter 1 Introduction
The main purpose of this book is to teach you how to construct math
ematical models of selected real-life problems. These problems were
chosen to include an ordinary differential equation in the model. To ob
tain a meaningful answer to the given problem, the differential equation
m ust be solved. The solution m ust then be interpreted in the light of
the assum ptions th at were made in the construction of the model, as
we have seen in §1.1.
A part from the differential equation, the m athem atical model will typ
ically also include prescribed values of y an d /o r the derivatives of y at
isolated points in the interval /, usually one or both of the endpoints.
points where the highest derivative does not exist. Let i(I) denote the
interior of 7 which means the largest open interval in 7. If 7 is itself
an open interval, then of course i(I) = 7; otherwise one or both the
endpoints will be excluded.
In this book, unless specifically stated otherwise, we shall use the word
“solution” in the following sense:
Note th a t the jum p in the value of f is the difference between the left
and right hand limits, which m ust both exist if the jum p is to be finite.
We shall continue the discussion of piecewise continuity in §2.8 and §4.2.
dy _ 2x ( s 2 + 1) _
dx~ *2 + 1 "
The function F ( x , y ) in (1.3.1) prescribes at each point ( x , y ) in the
plane where F is defined, a slope (or direction) for the solution curve
which passes through (x, y). We say briefly th a t the differential equation
prescribes a direction field. If we choose a suitable rectangular mesh in
the plane, a useful picture of the direction field can be obtained. At
each point (a, 6) of the mesh in the (#, y) —plane, a small line segment
(also called a lineal element) is drawn from (a, b) with slope F (a ,6 ).
Since each of these line segments is a tangent to a solution curve, the
picture of all these line segments gives an indication of the shape of the
solution curves. Obviously, the finer the rectangular mesh, the b etter
this indication will be - provided th a t the mesh is not finer than the
thickness of the line segments!
In this way an idea of the behaviour of the solution curves can be ob
tained, even though the solution of (1.3.1) may not be known.
Let us draw the direction field for the differential equation in (1.3.2)
on the subset R = {(x,t/) : —2 < x < 2, —2 < y < 2} with the
rectangular mesh points at ±2, ±1.5, ±1.0, ±0.5, and 0 for both x and
y. Then we need the slope at 81 mesh points. At the origin the slope is
1.3 Direction Fields 9
c/_ _ N (x >y)
F ( x , y ) = Tv r
D(x,y)
where N and D are defined (and finite) for all the points in the subset
of the plane under consideration.
Algorithm: DIRFIELD
(1) Specify the ranges [a, 6] on the x —axis and [c, d\ on the y —axis
(a, 6, c, and d are integers).
(8) Calculate F { x , y ) = 2 $ f $ .
(9) Draw the line segment from ( x , y ) with slope F ( x , y ) and length
h
2 *
(10) y = y + h.
(11) if y < rf, go to step 7; otherwise continue to the next step.
(12) x = x -f h.
(13) If x < 6, go to step 6; otherwise continue to the next step.
(14) End.
This algorithm m ust now be translated into a program for the com puter,
according to the com puter language of the reader’s choice. For example,
in TURBO.PASCAL 4.0 a program would look like this:
program D IR F IE L D ;
uses
Grap h;
va r
Grap hD river: integer;
GraphMode: integer;
j4 ,£ ,C ,D ,/ ,M ,./ V ,P ,Q ,S ,T , V, W: integer;
x ,y,h ,U ,L : real;
fun ction N U M (x ,y : real): real;
begin
N U M : = 2* x* (s* x + 1);
end;
fun ction D E N (x ,y: real): real;
begin
D E N := y
end;
begin
W r i t e l n ( ’D I R F I E L D ’);
W r i t e l n f 1E n t e r the intervals on the a x es :’) ;
W r i t e l n ( ’(Numbers m u s t be in teg ers !)’);
W r ite ln ( Left hand (n o n -po sitive) endpoint on x -a x is= : A — ’);
R e ad ( A );
W r i t e l n ( ’Right hand (no n -n ega tive) endpoint on x - a x i s = : B =
R e ad (B );
If A * A + B * B = 0 then W r i t e l n f ’ x -interva l zero!!’);
W r i t e l n ( ’L ower (no n -po sitive) endp oint on the y-axis: C = ’);
Read (C);
W r i t e l n f ’Upper (non-negative) endpoint on the y-axis: D = ’);
R e ad (D );
1.3 Direction Fields 11
The labels of the axes were om itted (on purpose to shorten the program ),
and the notation F in step 8 of the algorithm was not used in the
program. Note th at if the denom inator D is zero at a mesh point,
then the line segment is not drawn there. Hence the omission of a
line segment in the mesh implies th at the slope is undefined at th a t
particular mesh point.
Algorithm: OURGRAPH
(1) Specify the ranges [a, 6] on the x —axis and [c, d\ on the y —axis,
(a, 6, c, and d are integers.)
(5) Let x = a.
(7) x = x + h.
(9) End.
program O U R G R A P H ;
uses
Graph;
v ar
G r a p h D river: integer;
GraphMode: integer;
A , B t C , D , I ,M ,N ,P ,Q , S , T , V: integer;
x,y,h : real;
fu n ctio n F U N C (x : real): real;
begin
F U N C : = S Q R T (2 * ( c o s (x ) - h i ) ) ;
end;
begin
W r i t e l n ( ’0 U R G R A P H ’);
W r i t e l n ( ’E n te r the interva ls on the a xes:’);
W r i t e l n ( ’(N umbe rs m u st be in teg ers !)’);
W r i t e l n ( ’Left hand (n on -positive) endpoint on x-axis: A = ’);
R e ad(A );
1.4 Finding Zeros 13
In many of the applications in the rest of the book the following problem
crops up: find the zeros of the function f ( x ) on the interval [a, 6]. In
other words, find the numbers &2, • • • > such th a t /(& ,) = 0 and
a ^ h < b for i = 1 ,2 ,... ,n . If the problem cannot be solved by an
analytical m ethod, we m ust fall back on a numerical procedure. In this
section we briefly discuss two such procedures.
First of all we m ust determ ine a subinterval of [a, b] in which a specific
zero x = k lies. This is usually done by drawing a qualitative graph (in
14 C hapter 1 In tro d u ctio n
Figure 1.4.1: S e c a n t m e th o d
S e c a n t M e th o d
V - f ( xi ) x\
(1.4.1)
/( * o )- /(z i) * 0 - * l
We then repeat the process with x \ and x<i (instead of xq and X\) to
obtain x3, etc. In general with xn_i and x ni we obtain
* „+ i = * n - / (* » )/ (* * ) _ / (* B
X_ , ) > n = 2 ,3 ,... (1.4.2)
Algorithm: SECANT
(4) Set 1 = 2.
(5) Calculate X = x\ — / i f y - j l ) *
(6) P rint X.
(7) If \ X — x\\ < E then stop; otherwise continue to the next step.
(8) If I = N -|- 1 then print failure message and stop; otherwise con
tinue to the next step.
(10) Go to step 5.
(11) END.
e ' * - x = 0. (1.4.4)
16 Chapter 1 Introduction
program S E C A N T ;
v ar
I,N: integer;
X 0 ,X 1 ,E ,F 0 ,F 1 ,X:real;
f u n ctio n F(x: real).real;
begin
F := EXP(-x) - x
end;
begin
W r i t e l n ( ' S E C A N T M E T H O D 9);
W r ite ln ( ’The tolerance i s ’);
W r i t e l n ( ’E = ’);
R e ad (E );
W r i t e l n ( ’The m a xim u m number of itera tions allowed i s ’);
W r i t e l n ( ’N = ’);
R e a d (N );
W r i t e l n ( ’The f irst two approxim ation s a r e ’);
W r i t e l n ( ’XO = ’);
Read(X O);
W r i t e l n ( ’X l = ’);
r e a d ( X l );
W r i t e l n f ’The successive a ppro xim ations a r e ’);
F 0:= F (X0);
F l : = F ( X 1);
I : = 2;
While I < = N + 1 do
begin
X : = X I - F 1 *(X 1 - X 0 ) / ( F 1 - FO);
W r i t e l n ( ’x = ’,X );
If A B S ( X - X I ) < E then halt;
X 0:= XI;
F 0 : = FI;
X l : = X;
F1:=F(X);
I:= I + l;
end;
Writeln;
W r i t e l n ( ’Required accuracy not obtained!!!!’);
W r i t e l n ( ’E ith er try o th er initia l approxim ations XO and X I ’);
W r i t e l n ( ’or increase N a n d / o r E ’);
end. (Seca nt)
\ k - x n^ \ ^ C \ k - x n \P (1.4.5)
where p = and
_ \r m
~ 2 |/ '( * ) |
Another random document with
no related content on Scribd:
commun acord et dit ensi: «Chiers sires, vous nous
requerés d’une cose moult pesans, et qui trop ou
25 temps avenir poroit touchier au pays de Flandres et
à nos hoirs. Voirs est que nous ne savons signeur au
jour d’ui ou monde de qui nous amenons tant le
pourfit et l’avancement que nous ferions de vous.
Mais ceste cose nous ne poons faire de nous tant
30 seulement, se toute la communaulté de Flandres entirement
ne s’i acorde. Si se retraira cescuns devers
sa ville, et remousterons ceste besongne generalment
[99] as hommes de nostre ville. Et là où la plus sainne
partie se vorra acorder, nous l’acorderons ossi. Et
serons chi arrière dedens un mois, et vous en responderons
si à point que vous en serés bien contens.»
5 Li rois d’Engleterre et d’Artevelle n’en peurent
adonc avoir aultre response; si le vosissent il
bien avoir plus brief, se il peuist estre, mès nennil.
Si respondi li rois: «A le bonne heure!» Ensi se
departi cilz parlemens, et retournèrent li consaulz
10 des bonnes villes en leurs lieus.
Or demora Jakemes d’Artevelle encores un petit
dalés le roy d’Engleterre, pour le cause de ce que li
rois se descouvroit à lui fiablement de ses besongnes.
Et il li prommetoit toutdis et asseguroit qu’il le feroit
15 venir à sen entente, mès non fist, si com vous orés
en avant recorder. Car il se dechut quant il demora
derrière, et qu’il ne revint à Gand ossi tost que li
bourgois qui avoient esté à l’Escluse au parlement
envoiiet, de par tout le corps de le ville.
20 Quant li consaulz de Gand fu retournés en l’absence
d’Artevelle, il fisent assambler grans et petis
ens ou Marchiet. Et là remoustra li plus sages d’yaus
tous par avis sur quel estat li parlemens avoit esté à
l’Escluse, et quel cose li rois d’Engleterre requeroit,
25 par l’ayde et information d’Artevelle. Dont commencièrent
toutes manières de gens à murmurer sus lui,
et ne lor vint mies bien à plaisance ceste requeste.
Et disent, se il plaisoit à Dieu, il ne seroient jà sceu
ne trouvé en tel desloyauté que de voloir deshireter
30 leur naturel signeur, pour ahireter un estragne. Et
se partirent tout dou Marchiet, ensi que mal content,
et en grant hayne sus d’Artevelle.
[100] Or regardés comment les coses aviennent: car s’il
fust là ossi bien premierement venus qu’il ala à Bruges
et à Ippre remoustrer et preecier le querelle dou
roy d’Engleterre, il leur euist tant dit d’unes et d’autres,
5 qu’il fuissent tout acordet à sen oppinion, ensi
que cil des dessus dittes villes estoient. Mais il s’affioit
tant en se prospérité et grandeur, que il y pensoit
bien à retourner assés à temps. Quant il eut fait
son tour, il revint à Gand et entra en le ville et toute
10 se route, ensi que à heure de miedi. Chil de le ville,
qui bien savoient se revenue, estoient assamblé sus
le rue par où il devoit chevaucier à son hostel. Si
tost qu’il le veirent, il commencièrent à murmurer
et à bouter trois tiestes en un caperon, et à dire:
15 «Vechi cesti qui est trop grans mestres et qui voet
ordonner de le conté de Flandres à se volenté: ce
ne fait mies à souffrir.» Encores avoech tout ce, ou
avoit semet parolles parmi le ville que le grant tresor
de Flandres, que Jakemes d’Artevelle avoit assamblé
20 par l’espasse de neuf ans et plus qu’il avoit eu le regimen
et le gouvrenement de Flandres (car des rentes
dou conte il n’aleuoit nulles, mès les mettoit et avoit
mises toutdis arrière et en depos; et tenoit son estat
et avoit tenu le terme dessus dit sus l’amende des
25 fourfaitures de Flandres tant seulement); ce grant
tresor où il avoit denierz sans nombre, il l’avoit envoiiet
secretement en Engleterre. Ce fu une cose qui
moult engrigni et enflama chiaus de Gand.
Ensi que Jakes d’Artevelle chevauçoit parmi le
30 rue, il se perçut tantos qu’il y avoit aucune cose de
nouvel et contre lui. Car cil qui se soloient encliner
et oster leurs chaperons contre lui, li tournoient l’espaule
[101] et rentroient en leurs maisons; si se commença
à doubter. Car si tretost qu’il fu descendus
à son hostel, il fist fremer et hameder[318] portes et huis
et fenestres. A painnes eurent si varlet ce fait, quant
5 toute li rue où il demoroit fu toute couverte, devant
et derrière, de gens, et especiaument de menues gens
de mestier. Là fu ses hostelz environnés et assallis
devant et derrière, et rompus par force. Bien est
voirs que cil de laiens se deffendirent moult longement,
10 et en atierèrent et blecièrent pluiseurs; mès
finablement il ne peurent durer, car il estoient assalli
si roit que priès les trois pars de le ville estoient
à cel assaut. Quant Jakemars d’Artevelle vei l’effort,
et comment il estoit apressés, il vint à une fenestre
15 sus les rues, et se commença moult à humeliier et à
dire par trop biau langage, et à nu chief: «Bonne
gent, que vous fault? Qui vous muet? Pour quoi estes
vous si troublé sur moy? En quel manière vous puis
je avoir couroucié? Dittes le moy: je l’amenderai
20 plainnement à vostre volenté.» Donc respondirent
il tout à une vois, voire cil qui oy l’avoient: «Nous
volons avoir compte dou grant tresor de Flandres
que vous avés desvoiié sans nul title de raison.»
Donc respondi d’Artevelle moult doucement: «Certes,
25 signeur, ou tresor de Flandres ne pris je onques denier.
Or vous retraiiés bellement en vos maisons,
je vous en pri, et revenés chi demain au matin. Et
je serai si pourveus de vous faire et rendre bon
compte, que par raison il vous devera souffire.»
30 Donc respondirent il d’une vois: «Nennil, nennil!
[102] Nous le volons tantost avoir; vous ne nous escaperés
mies ensi. Nous savons de verité que vous l’avés
vuidiet de pieçà, et envoiiet en Engleterre, sans nostre
sceu: pour la quèle cause il vous fault morir.»
5 Quant d’Artevelle oy ce mot, il joindi ses mains
et commença à plorer moult tenrement, et dist:
«Signeur, tel que je sui vous m’avés fait, et me jurastes
jadis que contre tous hommes vous me deffenderiés
et garderiés, et maintenant vous me volés occire,
10 et sans raison! Faire le poés, se vous volés,
car je ne sui que uns seulz homs contre vous tous, à
point de deffense. Avisés vous pour Dieu, et retournés
au temps passé. Si considerés les grasces et les
grans courtoisies que de jadis vous ay faites. Vous
15 me volés rendre petit guerredon des grans biens que
dou temps passé je vous ay fais! Ne savés vous comment
toute marchandise estoit perie en ce pays: je
le vous recouvrai. En apriès, je vous ay gouvrené
en si grant pais que vous avés eu, le temps de mon
20 gouvrenement, toutes coses à volenté, blés, lainnes,
avoir et toutes marcheandises, dont vous estes recouvré
et en bon point.» Donc commencièrent il à criier
tout d’une vois: «Descendés, et ne nous sermonnés
plus de si hault, car nous volons avoir compte et
25 raison tantost dou grant tresor de Flandres que vous
avés gouvrené trop longement, sans rendre compte;
ce qu’il n’apertient mies à nul officiier qu’il reçoive
les biens d’un signeur et d’un pays, sans compter.»
Quant d’Artevelle vei que point ne se refroideroient
30 ne affreneroient, il recloy la fenestre, et s’avisa
qu’il wideroit par derrière, et s’en iroit en une eglise
qui joindoit priès de son hostel; mès ses hostelz estoit
[103] rompus et effondrés par derrière, et y avoit plus
de quatre cens personnes qui tout tiroient à lui
avoir. Finablement, il fu pris entre yaus, et là occis
sans merci; et li donna le cop de le mort uns teliers
5 qui s’appelloit Thumas Denis. Ensi fina d’Artevelle,
qui en son temps fu si grans mestres en Flandres.
Povres gens l’amontèrent premierement, et meschans
gens le tuèrent en le parfin. Ces nouvelles
s’espardirent tantost en pluiseurs lieus: si fu plains
10 des aucuns, et pluiseur en furent bien liet. A ce donc
se tenoit li contes Loeis de Flandres à Tenremonde.
Si fu moult joyans quant il oy dire que Jakemes
d’Artevelle estoit occis, car il li avoit estet moult
contraires en toutes ses besongnes. Non obstant ce,
15 ne s’osa il encores affiier sus chiaus de Flandres,
pour revenir en le ville de Gand.